Investing in Michigan s Future: Meeting the Early Childhood Challenge

Transcription

1 u u u u u u u u u u u u u u u u u u Investing in Michigan s Future: Meeting the Early Childhood Challenge

2 u u u u u u u u u u u u u u u u u u Investing in Michigan s Future: Meeting the Early Childhood Challenge Lawrence J. Schweinhart President High/Scope Educational Research Foundation Rachel Fulcher-Dawson Graduate Research Associate The Education Policy Center Michigan State University October 2006

3

4 Meeting the Early Childhood Challenge EXECUTIVE SUMMARY The experiences children have before the age of 5 can profoundly affect the rest of their lives. Many Michigan children are growing up in conditions of toxic stress that put them at a huge initial disadvantage when they begin school. The right early childhood care and education programs can help even the neediest children overcome these disadvantages. The state and the nation have a compelling interest in making sure that every child comes to school ready to succeed. This report presents results from research into the effects of early childhood programs on their participants, particularly low-income and at-risk children. The studies described offer the best evidence on the longterm effects of high quality early childhood programs. Such programs can have long-lasting positive effects on children and their families, whether measured by intellectual performance in childhood, school achievement in adolescence, reduced placements in special education, reduced retentions in grade, improved high school graduation rates, reduced arrest rates, or older age of mothers at the birth of their first child. These evaluations further suggest that the returns on investments in high-quality early childhood education are dramatically larger than the returns on almost any other public investment, with a return of anywhere from four to seventeen dollars for every dollar spent on programs. Studies of Head Start and state preschool programs report modest shortterm effects, including significant improvement in the literacy and social skills of children as well as in the behavior of their parents. The evidence described in this report strongly suggests that existing publicly funded preschool programs would benefit from becoming more like high-quality model preschool programs that have been proven effective if they are to achieve the long-term effectiveness of which they are capable. Michigan is prevented from reaping potentially large gains from investment in early childhood education because care and education for young children are provided under a patchwork quilt of policies, programs and providers at both the state and federal levels. Resources and responsibility are scattered across a diverse array of competing agencies, with insufficient attention given to the quality of programs, especially for the neediest children. Turning this patchwork of governance practices and programs into an efficient, effective system is a huge opportunity for Michigan policymakers but also a tremendous challenge. Providing high quality early childhood care and education to Michigan s young children, particularly those living in poverty and other at-risk conditions, is an investment that will pay substantial, tangible dividends down the road by raising skill levels across the state s workforce while reducing unemployment and welfare dependency, criminal justice costs and teen pregnancy rates. Many states are now moving toward comprehensive systems for early childhood i

5 Investing in Michigan s Future care and education and have adopted early childhood standards. Michigan is beginning to map out a comprehensive early childhood support system, but has yet to put significant resources behind this effort. The goal of Michigan s early childhood education policies in the immediate future should be a steady increase in the number of Michigan children participating in high-quality early childhood education programs. A strategy to achieve this goal requires action on five different fronts: Michigan must reduce the patchwork of programs and providers, rather than making it more complicated. Policy-makers must raise the floor in terms of the quality of care received by Michigan children by increasing standards and expectations for all early childhood education programs and providers. State policy-makers should use funds more effectively, shifting funds to higher-quality center-based programs such as the Michigan School Readiness Program and away from low-quality and custodial programs. Policy-makers should target resources toward the neediest children, including those living with toxic stress and those currently involved in low-quality early childhood care. Policy-makers should support efforts to develop and expand model programs for needy children in needy communities. Decisions about early childhood care and education made by state policymakers today will have profound and lasting effects on the economic and social well-being of children, their families and the state itself for decades to come. ii

6 Meeting the Early Childhood Challenge TABLE OF CONTENTS I. The Early Childhood Challenge II. Evidence of the Value of Highquality Early Childhood Care and Education III. Early Childhood Care and Education in Context IV. Early Childhood Care and Education in Michigan Today V. Michigan Early Childhood Policy Opportunities Notes iii

7 Investing in Michigan s Future iv

8 Meeting the Early Childhood Challenge LIST OF TABLES AND FIGURES Table 1 Characteristics of Three Long-Term Early Childhood Studies Table 2 Findings of Recent Short-Term Early Childhood Program Studies Figure 1 Nonparental Care and Education Arrangements for U.S. Children Table 3 U.S. Children under 6 years old, by type of weekly nonparental child care arrangements Table 4 Federal Early Childhood Programs Figure 2 Total CCDF Expenditures: Figure 3 U.S. Subsidized Child Care Providers, by Licensing and Type Table Michigan Subsidized Child Care Providers, by Licensing and Type Figure 4 Michigan Subsidized Providers, by Type Figure CCDF Funds in Michigan, by Type of Regulation Figure 6 Regulated CCDF Providers in Michigan vs. U.S.: v

9 Investing in Michigan s Future vi

10 Meeting the Early Childhood Challenge I THE EARLY CHILDHOOD CHALLENGE Two very different kinds of research into early childhood development now provide convincing evidence that converges on a common and common sense conclusion: the experiences children have before the age of five have profound influence upon the rest of their lives. 1 The first kind of research involves remarkable advances in the field of neuroscience, or brain research. The brain is a complex collection of interconnected cells that transmit information from one cell to another. The brain develops by building basic circuits between cells, which become the basis for more complex circuits, and so on. Thus, a child first learns to discriminate language sounds, then learns to combine these sounds into words, then sentences, and so forth. Genetics determines the timetable for brain development, but experience determines the actual construction of circuits between brain cells, known as brain architecture. A newborn baby has the capacity to learn any language in the world, but immediately begins to build the circuits that allow her to discriminate the sounds of her home language rather than other languages. Using sophisticated medical imaging techniques, such as positron emission topography and functional magnetic resonance imaging, researchers can actually see this brain architecture at work as the circuits fire during mental activity. The experiences children have before the age of five have profound influence upon the rest of their lives. While the early childhood experiences needed for healthy brain development are commonplace in the environments of most children, neuroscience has identified conditions that are dangerous to the developing brain, and from which young children should be protected. Young children cannot handle the toxic stress that arises from persistent, dysfunctional poverty in which they experience physical or emotional abuse or neglect. Toxic stress involves the absence of supportive relationships and harms the development of young children s brains. Simply put, the brains of young children are harmed by neglect, abuse and harsh environments. Neuroscience imaging shows that the brains of young children raised under such conditions are visibly less developed than the brains of young children raised under healthy conditions. Many Michigan children are growing up in conditions of toxic stress that put them at a huge initial disadvantage when they begin school. The state and the nation have a compelling interest in making sure that every child comes to school ready to succeed. 1 The conclusions expressed in the report are the author s own, and do not necessarily reflect the views of Michigan State University. We wish to express our thanks to Chris Reimann and Jeannie Patrick for their help with the editing and production of the report. 1

11 Investing in Michigan s Future The second kind of research involves rigorous, random-assignment studies of the effects of early childhood programs on children s development and subsequent success. Some of these studies focused specifically on model programs that is, programs designed as experimental models to test the efficacy of particular approaches to child care and development. These model programs were targeted at young children living in at-risk environments and provided children with respite from the toxic stress in their lives, as well as educational experiences that helped grow the circuitry of their brains. High-quality early childhood education can be expected to raise skill levels across the state s workforce and to reduce unemployment and welfare dependency, criminal justice costs and teen pregnancy rates. 2 The data produced by these studies provide solid evidence that high quality programs have significant, positive, long-lasting impacts on the participating children, their families and the state as a whole impacts that can be measured in higher test scores and graduation rates for the program participants, and in return-on-investment rates for policy-makers and taxpayers. In other words, the economic and social benefits of high quality early childhood care and education are significant. This is important news, because the numbers involved in early childhood care and education are sizeable. Michigan has more than 650,000 children under the age of five, almost 400,000 of whom have some kind of child care arrangement a number approximately equal to the total populations of Grand Rapids, Lansing and Troy combined. More than 80,000 child care providers across the state receive funding from one or more programs, from babysitting relatives and neighbors to child care centers and preschools staffed by early childhood specialists. There is also sizeable money involved in early childhood care and education. In fiscal 2003 the State of Michigan awarded nearly $500 million dollars in state and federal subsidies to child care providers and Michigan families paid hundreds of millions of dollars on their own. In all, the federal government spends more than $16 billion each year on early childhood programs. Unfortunately, Michigan is handicapped from reaping the maximum benefits of early childhood care and education by a patchwork of governance practices and programs that reflect a lack of clarity in the purpose and scope of early childhood policy. Some of these practices and programs were established long before the research on brain development and effective programs was available. Many of them have been shaped by their connection to other important policy domains such as welfare or school reform. Inevitably, all of them have been influenced by the ebb and flow of funding, as a succession of policy-makers has struggled to determine spending priorities in good economic times and bad. Turning this patchwork of governance practices and programs into an efficient, effective system is a huge opportunity for Michigan policymakers but also a tremendous challenge. Providing high quality early childhood care and education to Michigan s young children, particularly those living in poverty and other at-risk conditions, is an investment that

12 Meeting the Early Childhood Challenge will pay substantial, tangible dividends down the road by raising skill levels across the state s workforce while reducing unemployment and welfare dependency, criminal justice costs and teen pregnancy rates. Improving early childhood care and education in Michigan will require some tough decisions by policy-makers. However sound, an investment is a decision to allocate current money toward a future goal, making less money available to meet immediate wants or needs. More funding for early childhood care and education means higher taxes or less funding for other worthy goals. The number of children and families involved and the cost of bringing the quality of care and education up to the levels shown by the research to have significant lasting effects make universal access to high quality programs in Michigan a goal unlikely to be achieved in the near future, no matter how attractive the future payoffs might be. Then, too, there is the inertia of well-established programs and practices, all of which have stood the political test of time, but not all of which are supported by the data on early childhood program effectiveness. Improving the outcomes for young children individually and the state as a whole may mean reconsidering the value of continuing these programs and practices. This situation is complicated further by the fact that early childhood issues are both a state and federal concern, meaning the state must coordinate its spending and policies with Washington. Finally, all early childhood care and education takes place within a culture that traditionally reserves to parents the primary responsibilities and choices in the care and well-being of their children, while recognizing that taxpayers determine the level of financial and institutional support they are willing to provide. The challenge for policy-makers, then, is to make tough decisions about how much the state should invest in the future success of its youngest citizens and itself. For this reason, it is important for policy-makers to understand the evidence that supports these claims about early childhood care and education. The National Research Council and Institute of Medicine have prepared an excellent report on the research from neuroscience, From Neurons to Neighborhoods: the Science of Early Childhood Development, 1 so that evidence will not be reviewed here. This report presents results from the second strand of research, that on the effects of early childhood programs on their participants, particularly low-income and at-risk children. This report highlights research on the potential benefits of improving state policies concerning early childhood care and education. The health and well-being of Michigan s children require policy-makers to be thoughtful stewards in many other policy areas as well, such as health care, nutrition, environmental quality and product safety. 3

13 Investing in Michigan s Future Terminology Used in the Report In the presentation and discussion that follow, we use the term early childhood to refer to children from birth through the beginning of kindergarten, typically age 5. Infants are children from birth to 1 year old; toddlers are children 2 or 3 years old; preschoolers are children 3 and 4 years old, as well as those 5 year olds who have not yet started kindergarten. This paper defines early childhood care and education as those activities that involve spending time with children of these ages. Child care refers to any non-parental care arrangement; early childhood education adds school readiness and other child development activities to supervised care. Short-term early childhood studies look at program effects as far as first grade. Long-term early childhood studies look at program effects beyond the elementary school years, even into adulthood. This report uses the term poverty to mean income below the federal poverty level of $18,850 for a family of four and low-income to mean income below $37,700 for a family of four. These are the guidelines used by the U.S. Department of Health and Human Services and the Michigan Community Health Department. It was recently estimated that the basic budget for a family of four (defined as the cost of meeting basic needs for food and shelter) in the largest metropolitan areas of Michigan was about $35,

14 Meeting the Early Childhood Challenge II EVIDENCE OF THE VALUE OF HIGH-QUALITY EARLY CHILDHOOD CARE AND EDUCATION Although schooling traditionally begins at 5 or 6 years of age because most children begin to learn to read at this age, a great deal of learning takes place before traditional schooling begins learning that is even more fundamental. The abilities to listen and speak precede the abilities to read and write. Working vocabulary in speech and comprehension determines how good a reader a child becomes. The components of everyday problemsolving include the ability to categorize and rank, and to recognize and understand spatial, temporal and causal relationships. Being able to regulate one s own actions and resolve conflicts with other people are essential to living in a complex society. Such abilities, which can and ought to develop in early childhood, are important not only in themselves, but as doors to all the rest of learning. These are the curriculum goals of early childhood. Children exposed to them develop in ways that prepare them for school; those who are not do not and face challenges for the rest of their lives. Research on the effects of model preschool programs for young children living in poverty confirms the basic findings of brain research. Results from long-term studies of three experimental preschool programs the Carolina Abecedarian Study, the High/Scope Perry Preschool Study and the Chicago Child-Parent Center Study indicate that these high quality model programs have had strong, lasting effects on the children who participated in them. Research on the effects of model preschool programs for young children living in poverty confirms the basic findings of brain research. These experimental or model programs may have been particularly effective in part because of the extreme poverty of their participants. Nevertheless, the evidence from these programs strongly suggests that existing publicly funded preschool programs would benefit from becoming more like the model preschool programs if they are to achieve the longterm effectiveness of which they are capable. The results of short-term studies of the publicly funded preschool programs described below show that these programs all have modest effects on participants. In other words, these programs were found to contribute in meaningful ways to the development of the children who participated in them. Much of the research on early childhood is limited in two important ways. First, because of the intense nature of the data collection and analysis, many studies look at relatively small programs with limited numbers of participants over short periods of time. Second, because these programs 5

15 Investing in Michigan s Future are intended to help children in great need, providers are reluctant to randomize participation in the program essentially, to deny some children the potential benefit of participation by randomly assigning them to a non-program control group. These two characteristics of most studies limit the ability of researchers to make generalizations about the cause and extent of the benefits (or adverse effects) of program participation. At age 21, children who had participated in the early childhood program were one-third more likely to have graduated from high school or to have received a GED certificate and more than twice as likely to have attended a 4-year college. 6 The studies we review in this report do not share these limitations. Most involved random assignment of children to program or no-program groups, a research design that prevents selection bias and confers considerable confidence in the findings. The findings of the studies reported here that did not employ random assignment are consistent with the findings of the studies that did. Long-term Early Childhood Program Studies The long-term follow-up studies of three early childhood programs the Carolina Abecedarian Project study, the High/Scope Perry Preschool study and the Chicago Child-Parent Centers study stand out for their duration and methodological quality. These studies, summarized in Table 1, offer the best evidence of the long-term effects of high quality preschool programs. Craig Ramey and his colleagues at the University of North Carolina at Chapel Hill began the Carolina Abecedarian Study in Researchers randomly assigned 57 infants to a special program group and 54 infants to a typical child care group participating in the child care arrangements in homes and centers that were prevalent in the 1970s. The special program was a full-day, full-year day care program for children that lasted the five years from birth to elementary school. Some of the study participants also received follow-up support from kindergarten to grade 3. The program s goal was to enhance children s cognitive and personal characteristics so they would achieve greater school success. It offered infants and toddlers good physical care, optimal adult-child interaction, and a variety of playthings and opportunities to explore them. It offered preschoolers a developmentally appropriate preschool learning environment and provided training to parents in how to enact the activity-based curriculum at home. The Abecedarian study was the first to find benefits to participants of preschool programs in intellectual performance and academic achievement throughout their schooling. Mean IQs of children in the program group, which started out the same as those of the control group, were significantly higher during early childhood and remained higher through age 21. Mean achievement scores at age 15 ten years after program participation were 94 versus 88 in reading and 94 versus 87 in mathematics. By age 15, members of the program group were less likely to have been retained in grade or to have received special services. At age 21, children who had participated in the early childhood program were one-third more likely to have graduated from high school or to have received a GED certificate and

16 Meeting the Early Childhood Challenge more than twice as likely to have attended a four-year college. As teens, members of the program group were 40 percent less likely to have become parents; of those who were parents at 21, the average age at the birth of a first child was 19.1 for the program group and 17.7 for the no-program group. 4 Cost-benefit analysis of the Abecedarian program indicates that, in 2000 dollars discounted at 3 percent annually (converted from the 2002 dollars reported), the program cost $34,476 per child ($13,362 per child per year) and yielded benefits to society of $130,300 $3.78 return per dollar invested. 5 Most of the benefits came from mothers earnings while the children were enrolled in the program (54 percent), participants earnings as adults (28 percent), and health improvement due to less smoking (13 percent). The High/Scope Perry Preschool Study identified the short- and longterm effects of a high-quality preschool education program for young children living in poverty. 6 From 1962 through 1967, David Weikart and his colleagues in the Ypsilanti, Michigan, school district operated a preschool program for young children to help them avoid school failure and related problems. They identified a sample of 123 low-income African American children who were living in poverty and assessed to be at high risk of school failure. Researchers randomly assigned 58 of these children to a group that received two years of a high-quality preschool program at ages 3 and 4; the other 65 children received no preschool program. Because of the random assignment strategy, children s preschool experience is the best explanation for subsequent group differences in their performance over the years. Project staff collected data annually on both groups from ages 3 through 11 and again at ages 14, 15, 19, 27, and 40, with a missing data rate of only 6 percent across all measures. The Perry preschool program led to greater success for the participants in many aspects of their lives. The program group outperformed the no-program group on various intellectual and language tests from their preschool years up to age 7; on school achievement tests at ages 9, 10 and 14; and on literacy tests at ages 19 and 27. During their schooling, fewer program than no-program females were treated for mental impairment or retained in grade. More of the program group than the no-program group graduated from high school specifically, more program females than no-program females. At ages 15 and 19, the program group had better attitudes toward school than the no-program group. More of the program group than the no-program group were employed at 27 and at 40. Program participants had median annual earnings 20 percent higher than those in the control group at age 27 and 35 percent higher earnings at age 40. More of the program group than the no-program group owned their own homes at 27 and at 40. By the time they reached age 40, program participants had much lower arrest rates and were half as likely to have been sentenced to prison or jail (28 percent vs. 52 percent). 7

17 Investing in Michigan s Future Table 1 Characteristics of Three Long-Term Early Childhood Studies Characteristic Carolina High/Scope Chicago Child- Abecedarian Perry Parent Centers Design Beginning year Type of setting College town College town Major city Sample size ,539 Assignment to groups Random Random Existing classes Program purpose Research Research Service Program entry and exit age Program hours a day, days a week 8, 5 2½, 5 2½, 5 Program weeks a year, years 50, 5 35, 2 35, 2 Parent program Weekly home Family and visits health services School-age services Yes No Yes Control group experience Some child care No preschool No preschool arrangements program program Common Outcomes Intellectual performance tests Ages 3-21 Ages School achievement tests Age 15 Ages 7-27 Ages Placed in special education 25% vs. 48% 65% vs. 60% 14% vs. 25% Retained in grade 31% vs. 55% 35% vs. 40% 23% vs. 38% High school graduates 67% vs. 51% 65% vs. 45% 50% vs. 39% Males 50% vs. 54% 43% vs. 29% Females 84% vs. 32% 57% vs. 48% Arrested by 21 45% vs. 41% 15% vs. 25% 17% vs. 25% Age at birth of first child 19.1 vs vs Cost-benefit analysis a Program cost $34,476 $15,166 $6,956 Program cost per year $13,362 $8,540 $4,637 Public return, total $195,621 $26,637 Public return, per dollar invested $12.90 $3.83 Societal return, total $130,300 $258,888 $49,364 Societal return, per dollar invested $3.78 $17.07 $7.10 a All dollar entries are per participant in constant 2000 dollars discounted at 3 percent annually. 8

18 Meeting the Early Childhood Challenge Cost-benefit analysis indicates that, in constant 2000 dollars discounted at 3 percent, the economic return to society for the program was $258,888 per participant on an investment of $15,166 per participant $17.07 per dollar invested. Of that return, 76 percent of the benefits went to the general public and 24 percent went to program participants. Of the public return, 88 percent came from crime savings; the rest included education savings, increased taxes due to higher earnings, and welfare savings. Remarkably, 93 percent of the public return was due to the participation of males because of reductions in crime and incarceration rates. Beginning in 1985, the Chicago Longitudinal Study, conducted by Arthur Reynolds and his colleagues, examined the effects of the Chicago Child- Parent Centers (CPC) program offered by the nation s third-largest public school district. 7 This program was citywide and much larger in scale than the research programs of the High/Scope Perry Preschool and Abecedarian studies. Hence, the study sample was larger, with 1,539 low-income children (93 percent African American, 7 percent Hispanic) enrolled in 25 schools. In the sample, 989 had been in the CPC program and 550 had not. Children in this study went to their neighborhood schools and were not randomly assigned to groups. Preschool-program group members attended a part-day preschool program when they were 3 and 4 years old, while the no-preschool-program group did not. At age 5, some members of both groups attended part-day kindergarten programs, while others attended full-day kindergarten programs. The CPC program involved the agency s traditional family-support services and preschool education. Parent outreach was provided by a family-support coordinator and a parent-resource teacher. The classroom program emphasized attainment of academic skills through relatively structured learning experiences presented by the teacher. Reynolds found that the preschool-program group did significantly better than the no-preschool-program group in educational performance and social behavior. Participants experienced lower rates of grade retention and special education placement while in school. They also showed a higher rate of high school completion, almost half a year more of education, and a lower rate of juvenile arrests. Analysis of the costs and benefits of the Chicago Child-Parent Centers program indicates that, in 2000 dollars discounted at 3 percent annually (converted from the 1998 dollars reported), the program cost $6,956 per child (based on average participation of 18 months) and yielded benefits of $49,564 per participant, $7.10 return per dollar invested. 8 Benefits to the general public were $26,637 per participant, $3.83 per dollar invested, with the largest benefits coming from more taxes paid on higher earnings (28 percent), reduced crime victim costs (18 percent), and reduced costs of school remedial services (18 percent). The economic return to society for the program was $258,888 per participant on an investment of $15,166 per participant $17.07 per dollar invested. The preschoolprogram group did significantly better than the no-preschoolprogram group in educational performance and social behavior. 9

19 Investing in Michigan s Future Comparing the Long-term Studies The Abecedarian, High/Scope and Chicago studies differed in time and place. The High/Scope program operated in the 1960s when there were few if any other services offered in the community that the no-program group children might receive. The Abecedarian program operated in the 1970s and the Chicago program operated in the 1980s, when families made a variety of child care arrangements. Thus, the High/Scope study compares program experience to home experience while the Abecedarian and Chicago studies compare experience in an intensive program to typical child care experience. The Abecedarian and High/Scope studies were intentional studies from the beginning, involving random assignment of samples of 100-plus children, while the Chicago study evaluated an existing program and involved a sample of 1,500-plus children in preexisting classes with no scientific intervention in the enrollment process. One of the key common factors in the quality of these programs was the level of teacher qualifications. All three of these studies found that high quality early childhood programs can have long-lasting positive effects on children, whether measured by intellectual performance in childhood, school achievement in adolescence, reduced placements in special education, reduced retentions in grade, improved high school graduation rate, reduced arrest rates, or older age of mothers at the birth of their first child. One of the key common factors in the quality of these programs was the level of teacher qualifications. Finally, it should be noted that the studies of the High/Scope Perry, Abecedarian and Chicago Child-Parent Centers programs identified larger short-term effects than have recent short-term studies of other publicly funded programs described below. For example, children were found to have gained 4 points on the Peabody Picture Vocabulary Test (version III) during their Head Start year in the FACES study and in the five-state preschool study. On the other hand, children in the High/Scope Perry Preschool program gained 8 points on this test in their first year twice as much as one year of Head Start or state preschool programs and a total of 15 points in two years, nearly four times as much as one year of Head Start or state preschool programs. 9 These would be worthwhile outcomes for any program. But all three longitudinal studies also found economic returns that were at least several times as great as the initial program investment. For every dollar spent on the program, the Perry program returned $17, the Chicago program returned $7 per dollar invested and the Abecedarian program returned almost $4. Although it was not calculated for the Abecedarian program, public return (to taxpayers) constituted about two-thirds of the Perry program return and about half of the Chicago program return. The Perry program generated the highest return even though its cost fell midway between the other two programs. According to leading economists, including Nobel Laureate James Heckman, this evidence of large public returns on investment in high-quality early childhood programs is stronger than the evidence on returns for most other public investments. 10

20 Meeting the Early Childhood Challenge Recent Short-term Early Childhood Program Studies Several recent studies have looked at the short-term effects of Head Start and other federally funded early childhood programs. Table 2 summarizes the design and findings of these studies. Two of them looked at typical Head Start programs and two looked at enhanced Head Start programs. There were two evaluations of the Even Start program, a five-state preschool study, and 13 state preschool evaluations. Two studies looked specifically at the issue of quality at child care centers. Most of these studies randomly assigned children to program or noprogram groups. While the Comprehensive Child Development Program evaluation found no effects, the others found modest positive effects on children s literacy and social skills and parental behavior. The consensus finding of the short-term studies is that typical publicly funded early childhood programs produced significant benefits for the literacy and social skills of children as well as for the behavior of their parents. The Head Start Impact Study, now under way, involves a nationally representative sample of Head Start programs and the random assignment of children to Head Start or a control group. This study has so far provided results for entering 3-year-olds and entering 4-year-olds after one year in Head Start and will follow them through the end of kindergarten and first grade. 10 It has found evidence of small to moderate Head Start effects on children s literacy skills (pre-reading, pre-writing, parent-reported literacy skills, 3-year-olds vocabulary), reduced problem behaviors of 3-year-olds, increased access to health care, greater incidence of parents reading to their children, and reduced use of physical discipline of 3-year-olds. The consensus finding of the short-term studies is that typical publicly funded early childhood programs produced significant benefits for children s literacy and social skills. The Head Start Family and Child Experiences Survey (FACES) is a study of a representative national sample of Head Start programs in the U.S. 11 The first cohort of 3,200 children entered Head Start in fall 1997; the second cohort of 2,800 children entered Head Start in fall While in Head Start, children improved on important aspects of school readiness, narrowing the gap between them and the general population, though they still lagged behind. Relative to national norms, children made significant gains during their Head Start year, particularly in vocabulary and early writing skills. Children in Head Start grew in social skills and displayed less hyperactive behavior, especially if they started out more shy, aggressive or hyperactive. The study found that Head Start classrooms were of good quality. Most programs used a specific integrated curriculum. Use of these curricula and higher teacher salaries were predictive of positive child outcomes. Teachers educational credentials were linked to greater gains in early writing skills. In addition, provision of preschool services for a longer period each day was tied to greater cognitive gains by children. Based on follow-up of the 1997 cohort, Head Start graduates showed further progress toward national averages during kindergarten, with substantial gains in vocabulary, early mathematics, and early writing skills. Most Head Start graduates could identify most or all of the letters 11

21 Investing in Michigan s Future Table 2 Findings of Recent Short-Term Early Childhood Program Studies Study Design Findings Head Start Impact Study Head Start Family and Child Experiences Survey Head Start Comprehensive Child Development Program Evaluation Early Head Start Evaluation Even Start Evaluations Five-State Preschool Study State Preschool Evaluations NICHD Early Child Care Study Cost, Quality and Child Outcomes in Child Care Centers Study Nationally representative sample of about 5,000 children randomly assigned to Head Start or no Head Start and followed 1 year to date 2 cohorts of 3,200 and 2,800 Head Start children followed through kindergarten Sample of 4,410 children randomly assigned to program or not; each family had a case manager Sample of about 3,000 children randomly assigned to Early Head Start as infants and toddlers or not 2 studies of children randomly assigned to Even Start or not Sample of 5,071 children divided between those who did and did not make the age cutoff for program entry 13 studies of state preschool programs Sample of 1,364 infants in 1991 with 1,000 being followed up through age 15. Random sample of 100 child care centers in 4 states, non-profit and for-profit centers in four U.S. cities in the mid-1990s; 733 children followed up at age 7. Modest effects on children s literacy skills, reduced problem behavior, parent reading to children Modest gains in children s literacy and social skills in Head Start and kindergarten years No effects on child or parent outcomes Modest effects on children s cognitive, language, and socioemotional development through age 2; effects on parents behavior and child development knowledge One-year improvement in children s readiness skills; gains in adult literacy, GED certification, family support of children Improvements in children s vocabulary, print awareness skills, and early mathematics skills Modest effects on children s development, school performance, school attendance, grade retention Higher quality child care was associated with higher math, reading, and memory test scores through grade 3. Quality rated high in 24 percent of centers, medium in 65 percent, and low in 11 percent. Quality had a modest effect on children s cognitive and socioemotional development at age 7 12

22 Meeting the Early Childhood Challenge of the alphabet by the end of kindergarten, and more than half could recognize beginning sounds of words. An evaluation of some 3,000 infants and toddlers and their low-income families in the Early Head Start program, the federal program that began in 1995, found program effects through age When compared to a randomly assigned control group, Early Head Start children did better in modestly but statistically significant ways on measures of cognitive, language, and social-emotional development, and their parents scored significantly better than control-group parents on measures of parenting behavior and knowledge of infant-toddler development. The Head Start Comprehensive Child Development Program Evaluation randomly assigned 4,410 children and families living in poverty at 21 sites either to this program or no program and followed them for five years. 13 The program s comprehensive services centered on assigning a case manager to each family to help them meet their needs, but only 58 percent of the program group actually met with a case manager, as did 18 percent of the control group due to other programs. The study found no statistically significant, positive group differences on either child or parent outcomes, suggesting that families do not really profit from case management associated with early childhood programs. Two evaluations of the Even Start Family Literacy program randomly assigned children and families to Even Start or not. 14 Somewhat greater percentages of the Even Start group than the control group received various services, with 95 percent versus 60 percent participating in early childhood education, for example. Consequently, both groups experienced gains, with the Even Start group experiencing some greater gains, in adult literacy, adult GED attainment (22 percent vs. 6 percent in one of the studies), cognitive stimulation and emotional support by the family, and children s vocabulary. Even Start children improved their basic school readiness skills (e.g., recognition of colors, shapes, and sizes), but their non-even Start peers caught up with them a year later. Lack of compliance with group assignment may have led to underestimation of program effects. Researchers Steven Barnett, Cynthia Lamy and Kwanghee Jung led a study of the effects of five state-funded preschool programs on the academic skills of entering kindergartners. 15 It involved 5,071 children from Michigan, New Jersey, Oklahoma, South Carolina, and West Virginia. Children who made the age cutoff for program entry were compared to those who missed the age cutoff using a sophisticated research design (regression discontinuity). The programs were found to have statistically significant, meaningful effects on children s vocabulary, print awareness skills, and early mathematics skills. Nearly all the teachers working in the study programs had a four-year college degree with early childhood specialization. 13

23 Investing in Michigan s Future In 2001, Walter S. Gilliam and Edward F. Zigler reported that 13 of the 33 state preschool programs had received evaluations as of They summarize these evaluations as finding modest support for positive program effects on children s developmental performance, school performance and attendance, and reduced percentages of children held back a grade. The NICHD Early Child Care Study is a longitudinal study initiated by the National Institute of Child Health and Human Development (NICHD) in 1989 to look at the relationship between children s child care experiences and their developmental outcomes. 17 The sample began with 1,364 infants in 1991 and is continuing to follow up on 1,000 of them through age 15. The study team has involved researchers at several dozen universities. The study found that higher quality child care was associated with higher test scores on mathematics and reading achievement and with improved memory through third grade. Several studies of typical child care programs in the U.S. bear out the idea that high-quality programs contribute to children s development while low-quality programs do not. The Cost, Quality and Child Outcomes in Child Care Centers Study was a longitudinal study of how children s experience in center-based care and school related to their socio-emotional and cognitive outcomes concurrently and at age The study focused on a random sample of 100 non-profit and for-profit centers in four states in the mid-1990s. Observers found that 65 percent of centers posted a medium score on the Early Childhood Environment Rating Scale, percent scored high, and 11 percent scored low. A follow-up study examined 733 children from these settings from ages 4 to 8 as a function of their child care center experience, after controlling for their background characteristics. The findings indicate that center quality had a modest longterm effect on children s cognitive and socioemotional development. 14 We conclude this discussion of the evidence from the research by noting that long-term studies in a related field have employed random assignment of study participants to high-quality programs, long-term follow-up, and cost-benefit analysis to examine the effectiveness of a different type of program prenatal and infancy home visitation by nurses. 20 David Olds and his colleagues studied 400 children, 89 percent of them White, in semirural Elmira, New York. 21 They found 79 percent fewer verified reports of child abuse or neglect, 31 percent fewer subsequent births among participating mothers, a longer interval to the birth of the next child, fewer months receiving welfare, fewer behavior problems due to alcohol and drug abuse, and 69 percent fewer arrests of the mothers. Cost-benefit analysis of those high-risk families revealed that the program cost $7,208 per family (in 2000 dollars) and led to benefits of $29,262 per family, four times as much. 22 A similar study in urban Memphis, Tennessee, involved 1,139 pregnancies and 743 children; 92 percent were African-American and 98 percent of the mothers were unmarried. 23 It found that the nurse-visited mothers provided better care for their children and had fewer subsequent pregnancies, and their children were hospitalized for fewer days with injuries indicative of child abuse and neglect.

24 Meeting the Early Childhood Challenge These and other studies have found that the best programs use professional staff and target families with certain characteristics. They have also found that most home visit and other family support programs have little or no long-term effects of practical value. 24 Summary The research on the long-term effects of model early childhood programs on poor and low-income children clearly shows that high-quality programs provide significant, positive, long-lasting benefits to the children, their families and the state that provide significant returns on investment. Studies of Head Start and state preschool programs report more modest short-term effects, some of which meeting state achievement standards and reducing grade retention rates can affect state and local budget policies directly and immediately. The success of early childhood programs is closely related to the quality of the professional staff. Given these findings, the question for policy-makers becomes not whether but where, how and how much to invest in early childhood programs. In the next section, we provide a description of the current investments the state and federal governments are making in these programs. Given these findings, the question for policy-makers becomes not whether but where, how and how much to invest in early childhood programs. 15

25 Investing in Michigan s Future There are big challenges and big opportunities for Michigan policy-makers to influence the long-term health of the state and its youngest citizens. III EARLY CHILDHOOD CARE AND EDUCATION IN CONTEXT Neuroscience shows that early childhood experiences fundamentally affect the brain s development. The best and latest research on early childhood programs shows that high quality programs have long-lasting positive effects that more than offset their costs. This research also shows that the kinds and degrees of benefits kids get are related to a program s quality. What implications do these findings have for the early childhood programs and policies currently operating in Michigan? In order to answer this question, we first need to make clear what early childhood care and education in Michigan looks like, beginning with an understanding of the patchwork of state and federal programs involved. This chapter provides a big picture description of the programs. Policy-makers may be surprised by the numbers of children affected, as well as by the number of programs and providers and the amount of funding involved. These numbers represent both big challenges and big opportunities for Michigan policy-makers to influence the long-term health of the state and its youngest citizens. Figure 1 Nonparental Care and Education Arrangements for U.S. Children Under 6 in Nonrelative Home Care 16% Relative Home Care 22% Center Care 33% No Arrangement 40% 0 5% 10% 15% 20% 25% 30% 35% 40% 45% Percentage of Children in This Type of Care Note Some children had more than one type of care. Therefore, the percentages total more than 16

26 Meeting the Early Childhood Challenge Table 3 U.S. children under 6 years old, by type of weekly nonparental child care arrangements and age: 2001 Number of Some Weekly Type of Weekly Nonparental Care No Weekly Children (in Nonparental Care and Education Arrangement: Nonparental Age Thousands) Arrangement Center 1 Nonrelative Relative Care Arrangement Total 20, year-olds year-olds 3, year-olds 3, year-olds 3, year-olds 3, Less than 1 3, Source: Mulligan et al., Note Some children had more than one type of care. Therefore, the percentages total more than 100%. 1 Day care centers, Head Start programs, preschools, prekindergartens, and other early childhood programs. 2 The population is defined as children under 6 years old who have not yet entered kindergarten, only about one-fourth of 5-year-olds. The National Household Education Survey, taken periodically by the National Center for Educational Statistics (NCES), presents a national picture of early childhood care and education. 25 Figure 1 and Table 3 present the key findings of the 2001 Early Childhood Program Participation Survey conducted by the NCES. According to the NCES Survey, in 2001 the U.S. had 20.2 million children under 6 years old who had not yet entered kindergarten. 60 percent of these children had some type of nonparental care and education arrangement at least weekly; the participation rate increased from 40 percent for infants to 79 percent for preschoolers, due to the increase in the center participation rate. 33 percent received care and education in a center, a proportion that grew steadily from 8 percent for infants to 65 percent for 4-yearolds. 16 percent received care and education from a non-relative in a home. 22 percent received care and education from a relative in a home. 26 For 62 percent of children under age 6, care and education remained all in the family, with either parental (40 percent) or relative care and education (22 percent). The percentage of young children having some non-parental care and education arrangement at least weekly increased steadily with household income, from 53 percent for those with income of $10,000 or less to 72 percent for those with income of more than $75,000. It similarly increased with the mother s level of education, from 43 percent of children of mothers with less than high school to 74 percent of children of mothers with a graduate or professional degree. It was highest for non-hispanic Blacks (73 percent), followed by non-hispanic Whites (60 percent) and Hispanics (48 percent). It was well over twice as high for children of For 62 percent of children under age 6, care and education remained all in the family, with either parental (40 percent) or relative care and education (22 percent). 17